Stop copy-pasting between AI models. Roundtable AI is a local MCP server that lets your primary AI assistant delegate tasks to specialized models like Gemini, Claude, Codex, and Cursor. Solve complex engineering problems in parallel, directly from your IDE. Key Features: - Context Continuity: Shared project context across all sub-agents - Parallel Execution: All agents work simultaneously - Model
Add this skill
npx mdskills install askbudi/roundtableWell-documented MCP server enabling parallel AI agent delegation with clear setup and examples
1# Roundtable AI MCP Server23[](https://badge.fury.io/py/roundtable-ai)4[](https://opensource.org/licenses/MIT)5[](https://www.python.org/downloads/)67[](cursor://anysphere.cursor-deeplink/mcp/install?name=roundtable-ai&config=eyJ0eXBlIjoic3RkaW8iLCJjb21tYW5kIjoidXZ4IiwiYXJncyI6WyJyb3VuZHRhYmxlLWFpQGxhdGVzdCJdLCJlbnYiOnsiQ0xJX01DUF9TVUJBR0VOVFMiOiJjb2RleCxjbGF1ZGUsY3Vyc29yLGdlbWluaSJ9fQo=)89Stop copy-pasting between AI models. Roundtable AI is a local MCP server that lets your primary AI assistant delegate tasks to specialized models like Gemini, Claude, Codex, and Cursor. Solve complex engineering problems in parallel, directly from your IDE.1011**Key Features:**12- **Context Continuity**: Shared project context across all sub-agents13- **Parallel Execution**: All agents work simultaneously14- **Model Specialization**: Right AI for each task (Gemini's 1M context, Claude's reasoning, Codex's implementation)15- **Zero Markup**: Uses your existing CLI tools and API subscriptions16- **26+ IDE Support**: Works with Claude Code, Cursor, VS Code, JetBrains, and more1718## Table of Contents1920- [Quick Start](#quick-start)21- [What is Roundtable AI](#what-is-roundtable-ai)22- [Technical Architecture](#technical-architecture)23- [Why Multi-Agent vs Single AI](#why-multi-agent-vs-single-ai)24- [Real-World Examples](#real-world-examples)25- [Installation](#installation)26- [IDE Integration](#ide-integration)27- [Advanced Configuration](#advanced-configuration)28- [Contributing](#contributing)29- [License](#license)3031## Quick Start3233```bash34# Install Roundtable AI35pip install roundtable-ai3637# Check available AI tools38roundtable-ai --check3940# Start with all available tools41roundtable-ai4243# Use specific assistants only44roundtable-ai --agents codex,claude45```4647**One-liner for Claude Code:**48```bash49claude mcp add roundtable-ai -- roundtable-ai --agents gemini,claude,codex,cursor50```5152**Try this multi-agent prompt in your IDE:**53```markdown54The user dashboard is randomly slow for enterprise customers.5556Use Gemini SubAgent to analyze frontend performance issues in the React components, especially expensive re-renders and inefficient data fetching.5758Use Codex SubAgent to examine the backend API endpoint for N+1 queries and database bottlenecks.5960Use Claude SubAgent to review the infrastructure logs and identify memory/CPU pressure during peak hours.61```6263## What is Roundtable AI6465Roundtable AI is a local Model Context Protocol (MCP) server that coordinates specialized AI sub-agents to solve complex engineering problems. Instead of manually switching between different AI tools, you delegate tasks from a single prompt in your IDE, and Roundtable manages the coordination, context sharing, and response synthesis.6667### Key Benefits6869- **Context Continuity**: The primary agent provides shared, rich context to all sub-agents70- **Parallel Execution**: All agents work simultaneously, drastically reducing wait time71- **Model Specialization**: Use the right AI for each task - Gemini's 1M context for analysis, Claude's reasoning for logic, Codex for implementation72- **No Extra Cost**: Uses your existing CLI tools and API subscriptions with zero markup73- **Single Interface**: One prompt, multiple specialized responses, automatically synthesized7475## Technical Architecture7677```text78 +----------------------------------+79 | Your IDE (VS Code, Cursor, etc.) |80 | (Primary AI Assistant) |81 +----------------+-----------------+82 |83 (1. User prompt with subagent delegation)84 |85 +----------------v-----------------+86 | Roundtable MCP Server |87 | (localhost) |88 +----------------+-----------------+89 |90 (2. Dispatches tasks to sub-agent CLIs in parallel)91 |92+--------------------v--------------------+93| |94| +-----------+ +-----------+ +-----------+ |95| | Gemini | | Claude | | Codex | |96| | (Analysis)| | (Logic) | | (Implement)| |97| +-----------+ +-----------+ +-----------+ |98| |99+--------------------^--------------------+100 |101 (3. Sub-agents execute using local tools,102 e.g., read_file, run_shell_command)103 |104 +----------------+-----------------+105 | Roundtable MCP Server |106 | (Aggregates & Synthesizes) |107 +----------------+-----------------+108 |109(4. Returns a single, synthesized response)110 |111 +----------------v-----------------+112 | Your IDE (Primary AI Assistant) |113 +----------------------------------+114```115116### How It Works1171181. **Context Continuity**: The initial prompt and relevant file/project context are packaged by the primary agent. The MCP server passes this "context bundle" to each sub-agent, ensuring all participants have the same ground truth without manual copy-pasting.1191202. **Model Specialization**: Use the right model for the job. Leverage Gemini's 1M context for codebase analysis, Claude's reasoning for logic and implementation, and Codex's proficiency for code generation and reviews, all in one workflow.1211223. **No Extra Cost**: Roundtable invokes the CLI tools you already have installed and configured. It uses your existing API keys and subscriptions. We add no markup. The cost is exactly what you would pay running the tools manually.123124## Why Multi-Agent vs Single AI125126Because manual context-switching is slow, error-prone, and prevents deep analysis.127128### The Multi-Tab Workflow ❌129130- Manually copy-paste code and context between different AI chats131- Each agent starts fresh, unaware of other conversations or files132- You wait for one agent to finish before starting the next133- You are responsible for merging disparate, often conflicting, advice134- High risk of pasting outdated code or incorrect context135136### The Roundtable Workflow ✅137138- Delegate tasks from a single prompt in your IDE139- The primary agent provides shared, rich context to all sub-agents140- All agents work in parallel, drastically reducing wait time141- The final output automatically synthesizes the best insights from each model142- The entire workflow is a single, deterministic, and repeatable command143144## Real-World Examples145146Each example includes real code, logs, and explicit delegation to specialized sub-agents. Copy the whole block and paste it into your IDE assistant.1471481) Multi-Stack Debugging — Virtual War Room for Production Issues149150```markdown151I'm debugging a critical production issue. The user sees a "Failed to load data" message.152153Here is the browser console output:154```json155{156 "timestamp": "2024-09-24T10:05:21.123Z",157 "level": "error",158 "message": "API request failed for /api/v1/user/profile",159 "error": {160 "status": 500,161 "statusText": "Internal Server Error"162 }163}164```165166Here is the backend server log:167```168ERROR: Exception in ASGI application169File "/app/services/user_service.py", line 42, in get_user_profile170 user_data = await db.fetch_one(query)171ValueError: Database connection is not available172```173174Use Gemini SubAgent to analyze the logs from both stacks, correlate the events, and form a hypothesis about the root cause.175Use Codex SubAgent to analyze the Python backend traceback and suggest a specific code fix for the database connection error.176Use Claude SubAgent to review the frontend error handling and recommend more resilient patterns.177Use Cursor SubAgent to search the codebase for other files that might have similar database connection issues.178179At the end, aggregate all findings into a single incident report with root cause analysis and prioritized fixes.180```1811822) Performance Optimization — API Latency & Database Query Tuning183184```markdown185Our checkout API p95 latency increased from 220ms to 780ms. Need optimization strategy.186187PostgreSQL slow query log:188```sql189-- Duration: 2455.112 ms190SELECT c.name, COUNT(o.id) AS total_orders, SUM(p.amount) AS revenue191FROM companies c, orders o, payments p192WHERE c.id = o.company_id193 AND o.id = p.order_id194 AND c.region = 'North America'195GROUP BY c.name196ORDER BY revenue DESC;197```198199EXPLAIN ANALYZE shows:200```201Seq Scan on orders (cost=0.00..52000.00 rows=100000)202 Filter: (status = 'completed')203 Rows Removed by Filter: 134,201204```205206Node.js hotspot from profiling:207```javascript208// 40% CPU time209orders.map(o => ({ ...o, json: JSON.stringify(o) }));210211// N+1 query problem212for (const id of orderIds) {213 await fetchInventory(id);214}215```216217Use Claude SubAgent to analyze the EXPLAIN plan and identify why the query is slow.218Use Codex SubAgent to rewrite the SQL with proper JOINs and suggest indexes.219Use Gemini SubAgent to fix the N+1 query problem with batch fetching.220Use Cursor SubAgent to find all instances of JSON.stringify in hot code paths.221222Aggregate findings into a performance optimization plan with measurable improvements.223```224225## Installation226227### Using pip (Standard)228229```bash230pip install roundtable-ai231```232233### Using UV/UVX (Recommended for faster installs)234235```bash236uvx roundtable-ai@latest237```238239## IDE Integration240241Roundtable AI supports 26+ MCP-compatible clients. Here are the top 7:242243### 1. Claude Code244245**Using pip:**246```bash247claude mcp add roundtable-ai -- roundtable-ai --agents gemini,claude,codex,cursor248```249250**Using UVX:**251```bash252claude mcp add roundtable-ai -- uvx roundtable-ai@latest --agents gemini,claude,codex,cursor253```254255### 2. Cursor256257**One-Click Install:**258259[](cursor://anysphere.cursor-deeplink/mcp/install?name=roundtable-ai&config=eyJ0eXBlIjoic3RkaW8iLCJjb21tYW5kIjoidXZ4IiwiYXJncyI6WyJyb3VuZHRhYmxlLWFpQGxhdGVzdCJdLCJlbnYiOnsiQ0xJX01DUF9TVUJBR0VOVFMiOiJjb2RleCxjbGF1ZGUsY3Vyc29yLGdlbWluaSJ9fQo=)260261Or use this direct link:262```263cursor://anysphere.cursor-deeplink/mcp/install?name=roundtable-ai&config=eyJ0eXBlIjoic3RkaW8iLCJjb21tYW5kIjoidXZ4IiwiYXJncyI6WyJyb3VuZHRhYmxlLWFpQGxhdGVzdCJdLCJlbnYiOnsiQ0xJX01DUF9TVUJBR0VOVFMiOiJjb2RleCxjbGF1ZGUsY3Vyc29yLGdlbWluaSJ9fQo=264```265266**Manual Installation:**267268**File:** `.cursor/mcp.json`269270**Using pip:**271```json272{273 "mcpServers": {274 "roundtable-ai": {275 "command": "roundtable-ai",276 "env": {277 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"278 }279 }280 }281}282```283284**Using UVX:**285```json286{287 "mcpServers": {288 "roundtable-ai": {289 "type": "stdio",290 "command": "uvx",291 "args": [292 "roundtable-ai@latest"293 ],294 "env": {295 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"296 }297 }298 }299}300```301302### 3. Claude Desktop303304**File:** `~/.config/claude_desktop_config.json`305306**Using pip:**307```json308{309 "mcpServers": {310 "roundtable-ai": {311 "command": "roundtable-ai",312 "env": {313 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"314 }315 }316 }317}318```319320**Using UVX:**321```json322{323 "mcpServers": {324 "roundtable-ai": {325 "command": "uvx",326 "args": ["roundtable-ai@latest"],327 "env": {328 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"329 }330 }331 }332}333```334335### 4. VS Code336337**Add to `settings.json`:**338339**Using pip:**340```json341{342 "mcp.servers": {343 "roundtable-ai": {344 "command": "roundtable-ai",345 "env": {346 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"347 }348 }349 }350}351```352353**Using UVX:**354```json355{356 "mcp.servers": {357 "roundtable-ai": {358 "command": "uvx",359 "args": ["roundtable-ai@latest"],360 "env": {361 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"362 }363 }364 }365}366```367368### 5. OpenAI Codex369370**File:** `~/.codex/config.toml`371372**Using pip:**373```toml374# IMPORTANT: the top-level key is 'mcp_servers' rather than 'mcpServers'.375[mcp_servers.roundtable-ai]376command = "roundtable-ai"377args = []378env = { "CLI_MCP_SUBAGENTS" = "codex,claude,cursor,gemini" }379```380381**Using UVX:**382```toml383# IMPORTANT: the top-level key is 'mcp_servers' rather than 'mcpServers'.384[mcp_servers.roundtable-ai]385command = "uvx"386args = ["roundtable-ai@latest"]387env = { "CLI_MCP_SUBAGENTS" = "codex,claude,cursor,gemini" }388```389390### 6. Windsurf391392**File:** `~/.codeium/windsurf/mcp_config.json`393394**Using pip:**395```json396{397 "mcpServers": {398 "roundtable-ai": {399 "command": "roundtable-ai",400 "env": {401 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"402 }403 }404 }405}406```407408**Using UVX:**409```json410{411 "mcpServers": {412 "roundtable-ai": {413 "command": "uvx",414 "args": ["roundtable-ai@latest"],415 "env": {416 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"417 }418 }419 }420}421```422423### 7. Gemini CLI424425**File:** `~/.gemini/settings.json`426427**Using pip:**428```json429{430 "mcpServers": {431 "roundtable-ai": {432 "command": "roundtable-ai",433 "env": {434 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"435 }436 }437 }438}439```440441**Using UVX:**442```json443{444 "mcpServers": {445 "roundtable-ai": {446 "command": "uvx",447 "args": ["roundtable-ai@latest"],448 "env": {449 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"450 }451 }452 }453}454```455456### Additional IDE Support457458Roundtable AI integrates with **26+ different IDEs and AI coding tools**:459460#### JetBrains IDEs (IntelliJ, PyCharm, WebStorm, etc.)461462**Settings Path**: `Settings > Tools > AI Assistant > Model Context Protocol (MCP)`463464```json465{466 "name": "roundtable-ai",467 "command": "roundtable-ai",468 "transport": "stdio",469 "env": {470 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"471 }472}473```474475#### GitHub Copilot476477```json478{479 "github.copilot.mcp.servers": {480 "roundtable-ai": {481 "command": "roundtable-ai",482 "env": {483 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"484 }485 }486 }487}488```489490---491492### 🖥️ Desktop IDEs493494<details>495<summary><b>JetBrains AI Assistant</b> - IntelliJ, PyCharm, WebStorm, etc.</summary>4964971. **Settings Path**: `Settings > Tools > AI Assistant > Model Context Protocol (MCP)`4982. **Add New Server**: Click "+" to add new MCP server4993. **Configuration**:500 ```json501 {502 "name": "roundtable-ai",503 "command": "roundtable-ai",504 "transport": "stdio",505 "env": {506 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",507 }508 }509 ```5104. **Apply & Restart**: Apply settings and restart the IDE511512</details>513514<details>515<summary><b>Visual Studio 2022</b> - Microsoft's Flagship IDE</summary>516517Create `mcp_config.json` in your project root:518```json519{520 "servers": {521 "roundtable-ai": {522 "command": "roundtable-ai",523 "transport": "stdio",524 "env": {525 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"526 }527 }528 }529}530```531532**Alternative**: Use Extensions > Manage Extensions > Search for "Roundtable AI"533534</details>535536<details>537<summary><b>Zed</b> - High-Performance Code Editor</summary>538539Add to `settings.json` (Cmd/Ctrl + ,):540```json541{542 "context_servers": {543 "roundtable-ai": {544 "command": "roundtable-ai",545 "env": {546 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"547 }548 }549 }550}551```552553**Extension Alternative**: Search for "Roundtable AI" in Zed Extensions554555</details>556557---558559### 💻 CLI Tools560561<details>562<summary><b>Gemini CLI</b> - Google's Gemini Command-Line Interface</summary>563564Edit `~/.gemini/settings.json`:565```json566{567 "mcpServers": {568 "roundtable-ai": {569 "command": "roundtable-ai",570 "env": {571 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"572 }573 }574 }575}576```577578</details>579580<details>581<summary><b>Rovo Dev CLI</b> - Atlassian's Development CLI</summary>582583Configure via `rovo config` command:584```bash585# Add MCP server586rovo mcp add roundtable-ai roundtable-ai587588# Verify589rovo mcp list590```591592**Manual configuration** in `~/.rovo/config.json`:593```json594{595 "mcpServers": {596 "roundtable-ai": {597 "command": "roundtable-ai",598 "env": {599 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"600 }601 }602 }603}604```605606</details>607608<details>609<summary><b>Amazon Q Developer CLI</b> - Amazon's AI Development Assistant</summary>610611Edit configuration in `~/.aws/q-developer/config.json`:612```json613{614 "mcpServers": {615 "roundtable-ai": {616 "command": "roundtable-ai",617 "env": {618 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"619 }620 }621 }622}623```624625</details>626627<details>628<summary><b>Crush</b> - Terminal-Based AI Assistant</summary>629630Create or edit `crush.json` in your project:631```json632{633 "mcp": {634 "roundtable-ai": {635 "command": "roundtable-ai",636 "transport": "stdio",637 "env": {638 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"639 }640 }641 }642}643```644645</details>646647<details>648<summary><b>Warp</b> - AI-Powered Terminal</summary>649650Configure via Warp settings:6511. **Open Settings**: Cmd/Ctrl + ,6522. **Navigate to**: Features > AI > MCP Servers6533. **Add Server**:654 ```json655 {656 "name": "roundtable-ai",657 "command": "roundtable-ai",658 "env": {659 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"660 }661 }662 ```663664</details>665666---667668### 🤖 AI Assistants669670<details>671<summary><b>Claude Desktop</b> - Anthropic's Desktop Application</summary>672673Edit `~/.config/claude_desktop_config.json`:674```json675{676 "mcpServers": {677 "roundtable-ai": {678 "command": "roundtable-ai",679 "env": {680 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"681 }682 }683 }684}685```686687</details>688689<details>690<summary><b>Cline</b> - AI Assistant Extension</summary>691692**One-Click Install**:6931. Open Cline MCP Server Marketplace6942. Search for "Roundtable AI"6953. Click "Install"696697**Manual Configuration** in `cline_mcp_settings.json`:698```json699{700 "mcpServers": {701 "roundtable-ai": {702 "command": "roundtable-ai",703 "env": {704 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"705 }706 }707 }708}709```710711</details>712713<details>714<summary><b>BoltAI</b> - AI Assistant Application</summary>7157161. **Open BoltAI Settings**7172. **Navigate to**: Plugins > MCP Servers7183. **Add New Server**:719 - Name: `roundtable-ai`720 - Command: `roundtable-ai`721 - Environment Variables:722 ```723 CLI_MCP_SUBAGENTS=codex,claude,cursor,gemini724 ```725726</details>727728<details>729<summary><b>Perplexity Desktop</b> - AI Search and Research Assistant</summary>730731Configure in Perplexity settings:732```json733{734 "mcpConfig": {735 "servers": {736 "roundtable-ai": {737 "command": "roundtable-ai",738 "env": {739 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"740 }741 }742 }743 }744}745```746747</details>748749<details>750<summary><b>Qodo Gen</b> - AI Code Generation and Analysis Tool</summary>751752Add to Qodo Gen configuration:753```json754{755 "mcp_servers": {756 "roundtable-ai": {757 "command": "roundtable-ai",758 "env": {759 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"760 }761 }762 }763}764```765766</details>767768---769770### 🛠️ Specialized Tools771772<details>773<summary><b>Opencode</b> - Open-Source AI Code Editor</summary>774775Add to `opencode_config.json`:776```json777{778 "mcpServers": {779 "roundtable-ai": {780 "command": "roundtable-ai",781 "env": {782 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"783 }784 }785 }786}787```788789</details>790791<details>792<summary><b>OpenAI Codex</b> - OpenAI's Code Generation Model Interface</summary>793794Edit `config.toml`:795```toml796[mcp_servers.roundtable-ai]797command = "roundtable-ai"798env = { CLI_MCP_SUBAGENTS = "codex,claude,cursor,gemini" }799```800801</details>802803<details>804<summary><b>Kiro</b> - AI Development Assistant</summary>805806Configure in `~/.kiro/config.json`:807```json808{809 "mcpServers": {810 "roundtable-ai": {811 "command": "roundtable-ai",812 "env": {813 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"814 }815 }816 }817}818```819820</details>821822<details>823<summary><b>Trae</b> - AI Development Environment</summary>824825Add to Trae workspace configuration:826```json827{828 "mcp": {829 "servers": {830 "roundtable-ai": {831 "command": "roundtable-ai",832 "env": {833 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"834 }835 }836 }837 }838}839```840841</details>842843<details>844<summary><b>LM Studio</b> - Local Language Model Interface</summary>845846**One-Click Install**:8471. Navigate to **Program > Install > Edit mcp.json**8482. Search for "Roundtable AI" in marketplace8493. Click "Install"850851**Manual Configuration**:852```json853{854 "mcpServers": {855 "roundtable-ai": {856 "command": "roundtable-ai",857 "env": {858 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"859 }860 }861 }862}863```864865</details>866867<details>868<summary><b>Zencoder</b> - AI-Powered Coding Assistant</summary>869870Configure via Zencoder settings panel:871```json872{873 "mcp_configuration": {874 "roundtable-ai": {875 "command": "roundtable-ai",876 "env": {877 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"878 }879 }880 }881}882```883884</details>885886<details>887<summary><b>Augment Code</b> - AI-Powered Code Completion</summary>888889Add to Augment Code workspace settings:890```json891{892 "mcpServers": {893 "roundtable-ai": {894 "command": "roundtable-ai",895 "transport": "stdio",896 "env": {897 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"898 }899 }900 }901}902```903904</details>905906<details>907<summary><b>Roo Code</b> - AI Development Environment</summary>908909Configure in Roo Code project settings:910```json911{912 "ai_assistants": {913 "mcp_servers": {914 "roundtable-ai": {915 "command": "roundtable-ai",916 "env": {917 "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"918 }919 }920 }921 }922}923```924925</details>926927---928929### 🔧 Configuration Tips930931#### Environment Variables Reference932```bash933# Specify which AI assistants to enable934CLI_MCP_SUBAGENTS="codex,claude,cursor,gemini"935936937# Enable debug logging938CLI_MCP_DEBUG=true939940# Override availability checking941CLI_MCP_IGNORE_AVAILABILITY=true942```943944#### Command Alternatives945All IDEs support these equivalent commands:946- `roundtable-ai` (primary command)947- `roundtable-mcp-server` (descriptive alias)948- `python -m roundtable_mcp_server` (Python module)949- `npx @roundtable/mcp-server` (NPM package - coming soon)950951#### Verification952After installation, verify the integration works:953954```bash955# Check server availability956roundtable-ai --check957958# Test connection (varies by IDE)959# Most IDEs will show "Roundtable AI" in their AI assistant panel960```961962#### Troubleshooting9631. **Server not found**: Ensure `roundtable-ai` is in your PATH9642. **Permission denied**: Run `chmod +x $(which roundtable-ai)`9653. **Config not loaded**: Check file paths and JSON syntax9664. **No AI tools detected**: Run `roundtable-ai --check` first967968## Available MCP Tools969970Once integrated, you get access to:971972### Availability Checks973- `check_codex_availability` - Verify Codex CLI status974- `check_claude_availability` - Verify Claude Code CLI status975- `check_cursor_availability` - Verify Cursor CLI status976- `check_gemini_availability` - Verify Gemini CLI status977978### Unified Task Execution979- `execute_codex_task` - Run coding tasks through Codex980- `execute_claude_task` - Run coding tasks through Claude Code981- `execute_cursor_task` - Run coding tasks through Cursor982- `execute_gemini_task` - Run coding tasks through Gemini983984## Advanced Configuration985986### Environment Variables987```bash988# Specify which assistants to enable989export CLI_MCP_SUBAGENTS="codex,gemini"990991992# Enable all tools regardless of availability993export CLI_MCP_IGNORE_AVAILABILITY=true994995# Enable debug logging996export CLI_MCP_DEBUG=true997```998999### Command Line Options10001001```bash1002roundtable-ai --help10031004Options:1005 --agents TEXT Comma-separated list of agents (gemini,claude,codex,cursor)1006 --check Check availability of all AI tools1007 --debug Enable debug logging1008 --version Show version information1009 --help Show this message and exit1010```10111012## Contributing101310141. Fork the repository10152. Create a feature branch (`git checkout -b feature/amazing-feature`)10163. Commit your changes (`git commit -m 'Add some amazing feature'`)10174. Push to the branch (`git push origin feature/amazing-feature`)10185. Open a Pull Request10191020## License10211022This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.10231024---10251026**Built for developers who value their time.** Stop context-switching between AI tools and start solving problems faster with coordinated multi-agent workflows.10271028For more examples, advanced usage patterns, and troubleshooting guides, visit our [GitHub repository](https://github.com/askbudi/roundtable).
Full transparency — inspect the skill content before installing.